Research On The Calibration Of AHRS Sensors

Micro-Electro-Mechanical Systems(MEMS)sensors have the advantages of small size,low cost,and low power consumption.In recent years,attitude and heading reference systems(AHRS)that based on MEMS sensors have been widely used.A typical AHRS consists of a three-axis magnetometer,a three-axis accelerometer,and a three-axis gyroscope.AHRS can provide reliable information of 3D attitude,and can be used in the fields of unmanned aerial vehicles,land vehicle navigation,mobile robots,virtual reality,etc.But AHRS is vulnerable to the errors of MEMS sensors.Aiming at the problem of error calibration for vector field sensors(including accelerometer and magnetometer)in AHRS,this paper mainly contributes in the following aspects.Firstly,the error model of tri-axial accelerometer and magnetometer is established according to the analysis of sensor error sources.Secondly,the RANSAC algorithm is proposed to solve the interference problem caused by the data of three-axial magnetometer and accelerometer.This method uses an iterative approach to estimate the parameters most suitable for the error model from a set of data containing error points.Experiments have proved that the RANSAC algorithm can effectively improve the heading errors in the case of interference.Thirdly,the shortage of ellipsoid fitting methods is analyzed.It cannot ensure the alignment between the sensor triad and the coordinate system of the host platform.The dual inner product method combined with ellipsoid fitting method is proposed to overcome this problem and achieve a complete calibration of the three-axis vector sensor.The simulation experiments verify the effectiveness of the algorithm.Finally,a method using a biaxial sensor measuring three-dimensional vector fields is proposed.By constructing a biaxial sensor error model and the using the property of gravitational acceleration value in a short time remains constant to compensate errors,the three-dimensional component of vector field can be calculated.